xref: /openbmc/linux/net/sched/sch_gred.c (revision 95e9fd10) 
1 /*
2  * net/sched/sch_gred.c	Generic Random Early Detection queue.
3  *
4  *
5  *              This program is free software; you can redistribute it and/or
6  *              modify it under the terms of the GNU General Public License
7  *              as published by the Free Software Foundation; either version
8  *              2 of the License, or (at your option) any later version.
9  *
10  * Authors:    J Hadi Salim (hadi@cyberus.ca) 1998-2002
11  *
12  *             991129: -  Bug fix with grio mode
13  *		       - a better sing. AvgQ mode with Grio(WRED)
14  *		       - A finer grained VQ dequeue based on sugestion
15  *		         from Ren Liu
16  *		       - More error checks
17  *
18  *  For all the glorious comments look at include/net/red.h
19  */
20 
21 #include <linux/slab.h>
22 #include <linux/module.h>
23 #include <linux/types.h>
24 #include <linux/kernel.h>
25 #include <linux/skbuff.h>
26 #include <net/pkt_sched.h>
27 #include <net/red.h>
28 
29 #define GRED_DEF_PRIO (MAX_DPs / 2)
30 #define GRED_VQ_MASK (MAX_DPs - 1)
31 
32 struct gred_sched_data;
33 struct gred_sched;
34 
35 struct gred_sched_data {
36 	u32		limit;		/* HARD maximal queue length	*/
37 	u32		DP;		/* the drop parameters */
38 	u32		bytesin;	/* bytes seen on virtualQ so far*/
39 	u32		packetsin;	/* packets seen on virtualQ so far*/
40 	u32		backlog;	/* bytes on the virtualQ */
41 	u8		prio;		/* the prio of this vq */
42 
43 	struct red_parms parms;
44 	struct red_vars  vars;
45 	struct red_stats stats;
46 };
47 
48 enum {
49 	GRED_WRED_MODE = 1,
50 	GRED_RIO_MODE,
51 };
52 
53 struct gred_sched {
54 	struct gred_sched_data *tab[MAX_DPs];
55 	unsigned long	flags;
56 	u32		red_flags;
57 	u32 		DPs;
58 	u32 		def;
59 	struct red_vars wred_set;
60 };
61 
62 static inline int gred_wred_mode(struct gred_sched *table)
63 {
64 	return test_bit(GRED_WRED_MODE, &table->flags);
65 }
66 
67 static inline void gred_enable_wred_mode(struct gred_sched *table)
68 {
69 	__set_bit(GRED_WRED_MODE, &table->flags);
70 }
71 
72 static inline void gred_disable_wred_mode(struct gred_sched *table)
73 {
74 	__clear_bit(GRED_WRED_MODE, &table->flags);
75 }
76 
77 static inline int gred_rio_mode(struct gred_sched *table)
78 {
79 	return test_bit(GRED_RIO_MODE, &table->flags);
80 }
81 
82 static inline void gred_enable_rio_mode(struct gred_sched *table)
83 {
84 	__set_bit(GRED_RIO_MODE, &table->flags);
85 }
86 
87 static inline void gred_disable_rio_mode(struct gred_sched *table)
88 {
89 	__clear_bit(GRED_RIO_MODE, &table->flags);
90 }
91 
92 static inline int gred_wred_mode_check(struct Qdisc *sch)
93 {
94 	struct gred_sched *table = qdisc_priv(sch);
95 	int i;
96 
97 	/* Really ugly O(n^2) but shouldn't be necessary too frequent. */
98 	for (i = 0; i < table->DPs; i++) {
99 		struct gred_sched_data *q = table->tab[i];
100 		int n;
101 
102 		if (q == NULL)
103 			continue;
104 
105 		for (n = 0; n < table->DPs; n++)
106 			if (table->tab[n] && table->tab[n] != q &&
107 			    table->tab[n]->prio == q->prio)
108 				return 1;
109 	}
110 
111 	return 0;
112 }
113 
114 static inline unsigned int gred_backlog(struct gred_sched *table,
115 					struct gred_sched_data *q,
116 					struct Qdisc *sch)
117 {
118 	if (gred_wred_mode(table))
119 		return sch->qstats.backlog;
120 	else
121 		return q->backlog;
122 }
123 
124 static inline u16 tc_index_to_dp(struct sk_buff *skb)
125 {
126 	return skb->tc_index & GRED_VQ_MASK;
127 }
128 
129 static inline void gred_load_wred_set(const struct gred_sched *table,
130 				      struct gred_sched_data *q)
131 {
132 	q->vars.qavg = table->wred_set.qavg;
133 	q->vars.qidlestart = table->wred_set.qidlestart;
134 }
135 
136 static inline void gred_store_wred_set(struct gred_sched *table,
137 				       struct gred_sched_data *q)
138 {
139 	table->wred_set.qavg = q->vars.qavg;
140 }
141 
142 static inline int gred_use_ecn(struct gred_sched *t)
143 {
144 	return t->red_flags & TC_RED_ECN;
145 }
146 
147 static inline int gred_use_harddrop(struct gred_sched *t)
148 {
149 	return t->red_flags & TC_RED_HARDDROP;
150 }
151 
152 static int gred_enqueue(struct sk_buff *skb, struct Qdisc *sch)
153 {
154 	struct gred_sched_data *q = NULL;
155 	struct gred_sched *t = qdisc_priv(sch);
156 	unsigned long qavg = 0;
157 	u16 dp = tc_index_to_dp(skb);
158 
159 	if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
160 		dp = t->def;
161 
162 		q = t->tab[dp];
163 		if (!q) {
164 			/* Pass through packets not assigned to a DP
165 			 * if no default DP has been configured. This
166 			 * allows for DP flows to be left untouched.
167 			 */
168 			if (skb_queue_len(&sch->q) < qdisc_dev(sch)->tx_queue_len)
169 				return qdisc_enqueue_tail(skb, sch);
170 			else
171 				goto drop;
172 		}
173 
174 		/* fix tc_index? --could be controversial but needed for
175 		   requeueing */
176 		skb->tc_index = (skb->tc_index & ~GRED_VQ_MASK) | dp;
177 	}
178 
179 	/* sum up all the qaves of prios <= to ours to get the new qave */
180 	if (!gred_wred_mode(t) && gred_rio_mode(t)) {
181 		int i;
182 
183 		for (i = 0; i < t->DPs; i++) {
184 			if (t->tab[i] && t->tab[i]->prio < q->prio &&
185 			    !red_is_idling(&t->tab[i]->vars))
186 				qavg += t->tab[i]->vars.qavg;
187 		}
188 
189 	}
190 
191 	q->packetsin++;
192 	q->bytesin += qdisc_pkt_len(skb);
193 
194 	if (gred_wred_mode(t))
195 		gred_load_wred_set(t, q);
196 
197 	q->vars.qavg = red_calc_qavg(&q->parms,
198 				     &q->vars,
199 				     gred_backlog(t, q, sch));
200 
201 	if (red_is_idling(&q->vars))
202 		red_end_of_idle_period(&q->vars);
203 
204 	if (gred_wred_mode(t))
205 		gred_store_wred_set(t, q);
206 
207 	switch (red_action(&q->parms, &q->vars, q->vars.qavg + qavg)) {
208 	case RED_DONT_MARK:
209 		break;
210 
211 	case RED_PROB_MARK:
212 		sch->qstats.overlimits++;
213 		if (!gred_use_ecn(t) || !INET_ECN_set_ce(skb)) {
214 			q->stats.prob_drop++;
215 			goto congestion_drop;
216 		}
217 
218 		q->stats.prob_mark++;
219 		break;
220 
221 	case RED_HARD_MARK:
222 		sch->qstats.overlimits++;
223 		if (gred_use_harddrop(t) || !gred_use_ecn(t) ||
224 		    !INET_ECN_set_ce(skb)) {
225 			q->stats.forced_drop++;
226 			goto congestion_drop;
227 		}
228 		q->stats.forced_mark++;
229 		break;
230 	}
231 
232 	if (q->backlog + qdisc_pkt_len(skb) <= q->limit) {
233 		q->backlog += qdisc_pkt_len(skb);
234 		return qdisc_enqueue_tail(skb, sch);
235 	}
236 
237 	q->stats.pdrop++;
238 drop:
239 	return qdisc_drop(skb, sch);
240 
241 congestion_drop:
242 	qdisc_drop(skb, sch);
243 	return NET_XMIT_CN;
244 }
245 
246 static struct sk_buff *gred_dequeue(struct Qdisc *sch)
247 {
248 	struct sk_buff *skb;
249 	struct gred_sched *t = qdisc_priv(sch);
250 
251 	skb = qdisc_dequeue_head(sch);
252 
253 	if (skb) {
254 		struct gred_sched_data *q;
255 		u16 dp = tc_index_to_dp(skb);
256 
257 		if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
258 			net_warn_ratelimited("GRED: Unable to relocate VQ 0x%x after dequeue, screwing up backlog\n",
259 					     tc_index_to_dp(skb));
260 		} else {
261 			q->backlog -= qdisc_pkt_len(skb);
262 
263 			if (!q->backlog && !gred_wred_mode(t))
264 				red_start_of_idle_period(&q->vars);
265 		}
266 
267 		return skb;
268 	}
269 
270 	if (gred_wred_mode(t) && !red_is_idling(&t->wred_set))
271 		red_start_of_idle_period(&t->wred_set);
272 
273 	return NULL;
274 }
275 
276 static unsigned int gred_drop(struct Qdisc *sch)
277 {
278 	struct sk_buff *skb;
279 	struct gred_sched *t = qdisc_priv(sch);
280 
281 	skb = qdisc_dequeue_tail(sch);
282 	if (skb) {
283 		unsigned int len = qdisc_pkt_len(skb);
284 		struct gred_sched_data *q;
285 		u16 dp = tc_index_to_dp(skb);
286 
287 		if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
288 			net_warn_ratelimited("GRED: Unable to relocate VQ 0x%x while dropping, screwing up backlog\n",
289 					     tc_index_to_dp(skb));
290 		} else {
291 			q->backlog -= len;
292 			q->stats.other++;
293 
294 			if (!q->backlog && !gred_wred_mode(t))
295 				red_start_of_idle_period(&q->vars);
296 		}
297 
298 		qdisc_drop(skb, sch);
299 		return len;
300 	}
301 
302 	if (gred_wred_mode(t) && !red_is_idling(&t->wred_set))
303 		red_start_of_idle_period(&t->wred_set);
304 
305 	return 0;
306 
307 }
308 
309 static void gred_reset(struct Qdisc *sch)
310 {
311 	int i;
312 	struct gred_sched *t = qdisc_priv(sch);
313 
314 	qdisc_reset_queue(sch);
315 
316 	for (i = 0; i < t->DPs; i++) {
317 		struct gred_sched_data *q = t->tab[i];
318 
319 		if (!q)
320 			continue;
321 
322 		red_restart(&q->vars);
323 		q->backlog = 0;
324 	}
325 }
326 
327 static inline void gred_destroy_vq(struct gred_sched_data *q)
328 {
329 	kfree(q);
330 }
331 
332 static inline int gred_change_table_def(struct Qdisc *sch, struct nlattr *dps)
333 {
334 	struct gred_sched *table = qdisc_priv(sch);
335 	struct tc_gred_sopt *sopt;
336 	int i;
337 
338 	if (dps == NULL)
339 		return -EINVAL;
340 
341 	sopt = nla_data(dps);
342 
343 	if (sopt->DPs > MAX_DPs || sopt->DPs == 0 || sopt->def_DP >= sopt->DPs)
344 		return -EINVAL;
345 
346 	sch_tree_lock(sch);
347 	table->DPs = sopt->DPs;
348 	table->def = sopt->def_DP;
349 	table->red_flags = sopt->flags;
350 
351 	/*
352 	 * Every entry point to GRED is synchronized with the above code
353 	 * and the DP is checked against DPs, i.e. shadowed VQs can no
354 	 * longer be found so we can unlock right here.
355 	 */
356 	sch_tree_unlock(sch);
357 
358 	if (sopt->grio) {
359 		gred_enable_rio_mode(table);
360 		gred_disable_wred_mode(table);
361 		if (gred_wred_mode_check(sch))
362 			gred_enable_wred_mode(table);
363 	} else {
364 		gred_disable_rio_mode(table);
365 		gred_disable_wred_mode(table);
366 	}
367 
368 	for (i = table->DPs; i < MAX_DPs; i++) {
369 		if (table->tab[i]) {
370 			pr_warning("GRED: Warning: Destroying "
371 				   "shadowed VQ 0x%x\n", i);
372 			gred_destroy_vq(table->tab[i]);
373 			table->tab[i] = NULL;
374 		}
375 	}
376 
377 	return 0;
378 }
379 
380 static inline int gred_change_vq(struct Qdisc *sch, int dp,
381 				 struct tc_gred_qopt *ctl, int prio,
382 				 u8 *stab, u32 max_P,
383 				 struct gred_sched_data **prealloc)
384 {
385 	struct gred_sched *table = qdisc_priv(sch);
386 	struct gred_sched_data *q = table->tab[dp];
387 
388 	if (!q) {
389 		table->tab[dp] = q = *prealloc;
390 		*prealloc = NULL;
391 		if (!q)
392 			return -ENOMEM;
393 	}
394 
395 	q->DP = dp;
396 	q->prio = prio;
397 	q->limit = ctl->limit;
398 
399 	if (q->backlog == 0)
400 		red_end_of_idle_period(&q->vars);
401 
402 	red_set_parms(&q->parms,
403 		      ctl->qth_min, ctl->qth_max, ctl->Wlog, ctl->Plog,
404 		      ctl->Scell_log, stab, max_P);
405 	red_set_vars(&q->vars);
406 	return 0;
407 }
408 
409 static const struct nla_policy gred_policy[TCA_GRED_MAX + 1] = {
410 	[TCA_GRED_PARMS]	= { .len = sizeof(struct tc_gred_qopt) },
411 	[TCA_GRED_STAB]		= { .len = 256 },
412 	[TCA_GRED_DPS]		= { .len = sizeof(struct tc_gred_sopt) },
413 	[TCA_GRED_MAX_P]	= { .type = NLA_U32 },
414 };
415 
416 static int gred_change(struct Qdisc *sch, struct nlattr *opt)
417 {
418 	struct gred_sched *table = qdisc_priv(sch);
419 	struct tc_gred_qopt *ctl;
420 	struct nlattr *tb[TCA_GRED_MAX + 1];
421 	int err, prio = GRED_DEF_PRIO;
422 	u8 *stab;
423 	u32 max_P;
424 	struct gred_sched_data *prealloc;
425 
426 	if (opt == NULL)
427 		return -EINVAL;
428 
429 	err = nla_parse_nested(tb, TCA_GRED_MAX, opt, gred_policy);
430 	if (err < 0)
431 		return err;
432 
433 	if (tb[TCA_GRED_PARMS] == NULL && tb[TCA_GRED_STAB] == NULL)
434 		return gred_change_table_def(sch, opt);
435 
436 	if (tb[TCA_GRED_PARMS] == NULL ||
437 	    tb[TCA_GRED_STAB] == NULL)
438 		return -EINVAL;
439 
440 	max_P = tb[TCA_GRED_MAX_P] ? nla_get_u32(tb[TCA_GRED_MAX_P]) : 0;
441 
442 	err = -EINVAL;
443 	ctl = nla_data(tb[TCA_GRED_PARMS]);
444 	stab = nla_data(tb[TCA_GRED_STAB]);
445 
446 	if (ctl->DP >= table->DPs)
447 		goto errout;
448 
449 	if (gred_rio_mode(table)) {
450 		if (ctl->prio == 0) {
451 			int def_prio = GRED_DEF_PRIO;
452 
453 			if (table->tab[table->def])
454 				def_prio = table->tab[table->def]->prio;
455 
456 			printk(KERN_DEBUG "GRED: DP %u does not have a prio "
457 			       "setting default to %d\n", ctl->DP, def_prio);
458 
459 			prio = def_prio;
460 		} else
461 			prio = ctl->prio;
462 	}
463 
464 	prealloc = kzalloc(sizeof(*prealloc), GFP_KERNEL);
465 	sch_tree_lock(sch);
466 
467 	err = gred_change_vq(sch, ctl->DP, ctl, prio, stab, max_P, &prealloc);
468 	if (err < 0)
469 		goto errout_locked;
470 
471 	if (gred_rio_mode(table)) {
472 		gred_disable_wred_mode(table);
473 		if (gred_wred_mode_check(sch))
474 			gred_enable_wred_mode(table);
475 	}
476 
477 	err = 0;
478 
479 errout_locked:
480 	sch_tree_unlock(sch);
481 	kfree(prealloc);
482 errout:
483 	return err;
484 }
485 
486 static int gred_init(struct Qdisc *sch, struct nlattr *opt)
487 {
488 	struct nlattr *tb[TCA_GRED_MAX + 1];
489 	int err;
490 
491 	if (opt == NULL)
492 		return -EINVAL;
493 
494 	err = nla_parse_nested(tb, TCA_GRED_MAX, opt, gred_policy);
495 	if (err < 0)
496 		return err;
497 
498 	if (tb[TCA_GRED_PARMS] || tb[TCA_GRED_STAB])
499 		return -EINVAL;
500 
501 	return gred_change_table_def(sch, tb[TCA_GRED_DPS]);
502 }
503 
504 static int gred_dump(struct Qdisc *sch, struct sk_buff *skb)
505 {
506 	struct gred_sched *table = qdisc_priv(sch);
507 	struct nlattr *parms, *opts = NULL;
508 	int i;
509 	u32 max_p[MAX_DPs];
510 	struct tc_gred_sopt sopt = {
511 		.DPs	= table->DPs,
512 		.def_DP	= table->def,
513 		.grio	= gred_rio_mode(table),
514 		.flags	= table->red_flags,
515 	};
516 
517 	opts = nla_nest_start(skb, TCA_OPTIONS);
518 	if (opts == NULL)
519 		goto nla_put_failure;
520 	if (nla_put(skb, TCA_GRED_DPS, sizeof(sopt), &sopt))
521 		goto nla_put_failure;
522 
523 	for (i = 0; i < MAX_DPs; i++) {
524 		struct gred_sched_data *q = table->tab[i];
525 
526 		max_p[i] = q ? q->parms.max_P : 0;
527 	}
528 	if (nla_put(skb, TCA_GRED_MAX_P, sizeof(max_p), max_p))
529 		goto nla_put_failure;
530 
531 	parms = nla_nest_start(skb, TCA_GRED_PARMS);
532 	if (parms == NULL)
533 		goto nla_put_failure;
534 
535 	for (i = 0; i < MAX_DPs; i++) {
536 		struct gred_sched_data *q = table->tab[i];
537 		struct tc_gred_qopt opt;
538 
539 		memset(&opt, 0, sizeof(opt));
540 
541 		if (!q) {
542 			/* hack -- fix at some point with proper message
543 			   This is how we indicate to tc that there is no VQ
544 			   at this DP */
545 
546 			opt.DP = MAX_DPs + i;
547 			goto append_opt;
548 		}
549 
550 		opt.limit	= q->limit;
551 		opt.DP		= q->DP;
552 		opt.backlog	= q->backlog;
553 		opt.prio	= q->prio;
554 		opt.qth_min	= q->parms.qth_min >> q->parms.Wlog;
555 		opt.qth_max	= q->parms.qth_max >> q->parms.Wlog;
556 		opt.Wlog	= q->parms.Wlog;
557 		opt.Plog	= q->parms.Plog;
558 		opt.Scell_log	= q->parms.Scell_log;
559 		opt.other	= q->stats.other;
560 		opt.early	= q->stats.prob_drop;
561 		opt.forced	= q->stats.forced_drop;
562 		opt.pdrop	= q->stats.pdrop;
563 		opt.packets	= q->packetsin;
564 		opt.bytesin	= q->bytesin;
565 
566 		if (gred_wred_mode(table))
567 			gred_load_wred_set(table, q);
568 
569 		opt.qave = red_calc_qavg(&q->parms, &q->vars, q->vars.qavg);
570 
571 append_opt:
572 		if (nla_append(skb, sizeof(opt), &opt) < 0)
573 			goto nla_put_failure;
574 	}
575 
576 	nla_nest_end(skb, parms);
577 
578 	return nla_nest_end(skb, opts);
579 
580 nla_put_failure:
581 	nla_nest_cancel(skb, opts);
582 	return -EMSGSIZE;
583 }
584 
585 static void gred_destroy(struct Qdisc *sch)
586 {
587 	struct gred_sched *table = qdisc_priv(sch);
588 	int i;
589 
590 	for (i = 0; i < table->DPs; i++) {
591 		if (table->tab[i])
592 			gred_destroy_vq(table->tab[i]);
593 	}
594 }
595 
596 static struct Qdisc_ops gred_qdisc_ops __read_mostly = {
597 	.id		=	"gred",
598 	.priv_size	=	sizeof(struct gred_sched),
599 	.enqueue	=	gred_enqueue,
600 	.dequeue	=	gred_dequeue,
601 	.peek		=	qdisc_peek_head,
602 	.drop		=	gred_drop,
603 	.init		=	gred_init,
604 	.reset		=	gred_reset,
605 	.destroy	=	gred_destroy,
606 	.change		=	gred_change,
607 	.dump		=	gred_dump,
608 	.owner		=	THIS_MODULE,
609 };
610 
611 static int __init gred_module_init(void)
612 {
613 	return register_qdisc(&gred_qdisc_ops);
614 }
615 
616 static void __exit gred_module_exit(void)
617 {
618 	unregister_qdisc(&gred_qdisc_ops);
619 }
620 
621 module_init(gred_module_init)
622 module_exit(gred_module_exit)
623 
624 MODULE_LICENSE("GPL");
625